ResiDualGAN: Resize-Residual DualGAN for Cross-Domain Remote Sensing Images Semantic Segmentation

The performance of a semantic segmentation model for remote sensing (RS) images pre-trained on an annotated dataset greatly decreases when testing on another unannotated dataset because of the domain gap. Adversarial generative methods, e.g., DualGAN, are utilized for unpaired image-to-image translation to minimize the pixel-level domain gap, which is one of the common approaches for unsupervised domain adaptation (UDA). However, the existing image translation methods face two problems when performing RS image translation: (1) ignoring the scale discrepancy between two RS datasets, which greatly affects the accuracy performance of scale-invariant objects; (2) ignoring the characteristic of real-to-real translation of RS images, which brings an unstable factor for the training of the models. In this paper, ResiDualGAN is proposed for RS image translation, where an in-network resizer module is used for addressing the scale discrepancy of RS datasets and a residual connection is used for strengthening the stability of real-to-real images translation and improving the performance in cross-domain semantic segmentation tasks. Combined with an output space adaptation method, the proposed method greatly improves the accuracy performance on common benchmarks, which demonstrates the superiority and reliability of ResiDualGAN. At the end of the paper, a thorough discussion is conducted to provide a reasonable explanation for the improvement of ResiDualGAN. Our source code is also available.